Biomart issue, why so few 3'utrs?

Question

[RNAseqer]

I have previously used the biomart webportal to dow nload fastas for the 3'utrs of a gene-stable ensemble id list. Typically I limit my output to "MANE Select" as I am trying to get just one 3'utr for each gene. However, I recently provided a list of ~2000 ensemble ids and only got 3'utrs for only ~125 of these. I know there are more than 125 protein coding genes in this list. Has biomart recently undergone changes that could explain why I am getting so few 3'utrs? If the problem lies in there being relatively few 3'utrs classified by Mane Select is there another set of parameters I could use to get a non-redundant list ? If so, what criteria would be used to select a single 3'utr when there are multiple annotated for a given gene?

Any help would be appreciated.

[i.sudbery]

Which species is this?

[GenoMax]

AFAIK "MANE" project is only for human protein-coding genes.

One potential explanation. Out of the 2000 ID's only 125 may be MANE.

RNAseqer : You could post a few examples of ID's where you are and are not able to find UTR.

Answer 1

[i.sudbery]

The MANE-select transcripts are pairs of identically annotated transcripts in ReqSeq and Ensembl. The 3' UTR is the region where otherwise similar transcripts are likely to differ between Ensembl and RefSeq. I know that a lot of work has recently gone in to harmonising the 3' UTRs, but I don't know how far this has gone.